2-1-1 Analysis

Load Libraries

require(tidyverse)

## Loading required package: tidyverse

## -- Attaching packages --------------------------------------- tidyverse 1.3.1 --

## v ggplot2 3.3.5     v purrr   0.3.4
## v tibble  3.1.5     v dplyr   1.0.7
## v tidyr   1.1.4     v stringr 1.4.0
## v readr   2.0.2     v forcats 0.5.1

## -- Conflicts ------------------------------------------ tidyverse_conflicts() --
## x dplyr::filter() masks stats::filter()
## x dplyr::lag()    masks stats::lag()

require(lubridate)

## Loading required package: lubridate

## 
## Attaching package: 'lubridate'

## The following objects are masked from 'package:base':
## 
##     date, intersect, setdiff, union

require(stringr)
require(readxl)

## Loading required package: readxl

require(arsenal)

## Loading required package: arsenal

## 
## Attaching package: 'arsenal'

## The following object is masked from 'package:lubridate':
## 
##     is.Date

require(gtsummary)

## Loading required package: gtsummary

require(plotly)

## Loading required package: plotly

## 
## Attaching package: 'plotly'

## The following object is masked from 'package:ggplot2':
## 
##     last_plot

## The following object is masked from 'package:stats':
## 
##     filter

## The following object is masked from 'package:graphics':
## 
##     layout

needs <- read.csv("../Data/Cleaned/01_Needs.csv")
names(needs) <- toupper(names(needs))
referrals <- read.csv("../Data/Cleaned/01_referrals.csv")
names(referrals) <- toupper(names(referrals))

AS per discussions, Housing is the primary need for all clients, so needs variable becomes unnecessary to include. Removed Needs from the structure and keep unique values

needs <- needs %>% select(-c(NEED)) %>% distinct()

Q1 —-

5. Cross tab: CE Screened (all categories) filtered by Need=Housing x Gender Excluding DV - Referral, as it’s determined along with some other CE_Screened value

Calculating for DV referral across calls

7. Crosstab: CE Screened (all categories) filtered by Need=Housing x Disability Status

Calculating for DV referal across calls

1. Crosstab: CE Screened (all categories) filtered by Need=Housing x MILITARY_STATUS

Q2 -

Q2.1 – REFERAL - GENDER

q2 %>% select(GENDER,REFERRAL) %>% tbl_summary(by = GENDER,missing = "no")

## 381 observations missing `GENDER` have been removed. To include these observations, use `forcats::fct_explicit_na()` on `GENDER` column before passing to `tbl_summary()`.

Q2.2 – REFERAL - DISABILITY_STATUS

q2 %>% select(DISABILITY_STATUS,REFERRAL) %>% tbl_summary(by = DISABILITY_STATUS,missing = "no")

## 413 observations missing `DISABILITY_STATUS` have been removed. To include these observations, use `forcats::fct_explicit_na()` on `DISABILITY_STATUS` column before passing to `tbl_summary()`.

Q2.3 – REFERAL - MILITARY_STATUS

q2 %>% select(MILITARY_STATUS,REFERRAL) %>% tbl_summary(by = MILITARY_STATUS,missing = "no")

## 610 observations missing `MILITARY_STATUS` have been removed. To include these observations, use `forcats::fct_explicit_na()` on `MILITARY_STATUS` column before passing to `tbl_summary()`.

knitr::opts_chunk$set(echo = TRUE, class.source = "watch-out")

1 Analysis of 2-1-1 Calls

This analysis provides insights on the trends of 2-1-1 calls between the years 2019 and 2021.

The analyses uses the raw 2-1-1 call logs data and models the trends of key indicators

The data processing and modeling code is available as STATA do files in “S:_CoordinatedEntry\2021-coordinated-entry-analysisFiles”

The report below gives a general description of the measurement and the output figures

1.1 Measuring Call Statistics

THe raw data contained daily averages of call handle time, wait time, hold time between January 2019 and September 2021.

1.2 Aggregate Inbound and Callback times by date and Calculate Weighted Averages.

The dataset includes average times for inbound calls and callbacks in separate rows. The average for a given indicator, say, average call time is the average of both inbound and callback times. Since each of these two types is reported as a daily average with different volume of calls, we calculate the overall average of the two by weighting inbound and callback averages by the respective number of calls.

Formula - Weighted_calltime = avg_inbound##inboundcalls + avg_callback##callbackcalls) / total calls

1.3 Modeling the Trends in Key Indicators

Each indicator was plotted across time and a local polynomial smoothing trend was modeled to represent the moving average of the indicators across time. Following figures show the local averages that could be interpreted as average for specific periods within the time span of the data.

1.3.1 Calls - English

Average Calls - English

1.3.2 Calls - Spanish

### Agents - English

Agents - English

1.3.3 Agents - Spanish

Average Call Length - English

1.3.4 Call Length - English

Average Call Length - English

1.3.5 Call Length - Spanish

Average Call Length - English

1.3.6 Wait time - English

Average Wait time - English

1.3.7 Wait time - Spanish

Average Wait time - Spanish

1.3.8 Hold time - English

Average Hold time - English

1.3.9 Hold time - Spanish

Average Hold time - Spanish